REAV-RRT*: A Path Planning Algorithm with High Space Redundancy for Autonomous Robots in Dynamic Environments

In industries such as manufacturing and warehouse logistics, as labor costs rise, the demand for automation and intelligent solutions from enterprises is increasing. Autonomous mobile robots, as an important component of this, have received widespread attention. The path planning problems require robots to find a feasible path from the starting position to the target position in the working environment consisting of obstacles. This paper proposes Radial Extension Admissible Velocity Rapidly exploring Random Trees (REAV-RRT*) algorithm, which is aimed at generating a path with high space redundancy and path smoothness while controlling the consequent increase of processing time and path length. The Obstacle Edge Circles are assigned to obstacles after vertices identification or edge point down-sampling, and the process of radial extension is inserted to improve path space redundancy compared with RRT variants. With perceived information of dynamic obstacles, it applies Velocity Obstacle (VO) and Admissible Velocity Region to achieve obstacles avoidance via rolling angle adjustments. Generally, REAV-RRT* reduces computation time by
40% and improves path smoothness by 37% compared to standard RRT* and A* algorithms. Simulation results also confirm the capability of the proposed algorithm in achieving performance improvements in terms of space redundancy (42%) and path length (4%) while controlling the consequent increase of processing time and path length compared with Artificial Potential Fields (APF) and Simulated Annealing (SA).